Introduction

What my analysis is about

For my corpus I will look at the repertoire of Bring Me The Horizon. Since the release of their first four albums, which can, according to Google, all be classified as “metalcore”, their music has progressed more towards alternative-/pop-/electronic-rock and they even released an EP with experimental and electronic music. Having been in their fanbase since around 2013, when Sempiternal was released, I myself experienced the changes in their music. Some people argue that only their “old” music is good and their new music is “too soft”, leading to them not listening to the band’s music anymore. My question is, what exactly has changed about their music, which aspects have remained similar to their early music and which haven’t?

I definitely think that BMTH’s music after 2013 has changed significantly compared to their earlier music and it will be most interesting to dive deeper into their work from that time on. As for specific tracks to look at Pray for Plagues, Chelsea Smile and Shadow Moses from their earlier work and Drown, Mantra and Ludens from their more recent work might be representative tracks, as they were all released as singles and thus viewed as important. Some of their more recent albums include different genres within them, so it might also be useful to examine those more closely and determine what and how big those differences are and whether they are significant.

…………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………..

General

First things first: Different music styles over the years compared


For this plot I have split Bring Me The Horizon’s music into three different categories, “Old Music”, their first four albums, “New music”, their three most recent releases, with exception of their experimental music from Music to listen to~, which is the third category. This album isn’t exactly representative of either their “new” or “old” music style, so I made it a separate category (hence “Experimental”). It is easily visible that BMTH’s new music is much more varied than their old music. Considering that high energy + low valence = angry music, and high energy + high valence = happy music, we could conclude that their music has overall become happier. The experimental music has very low valence values, like BMTH’s old music, and is also more angry and negative. Each category seems to have outliers that fall into the ‘sad music’ area, they are labeled in the plot.

With the highest danceability score being 0.6 out of a possible maximum of 1, most of their music isn’t very danceable, but their newer music generally scores higher. The speechiness seems to be higher with higher valence scores, but generally the music doesn’t seem very speechy. Looking at all their work together, the music seems very energetic and seems to have progressed from negative and angry towards a more diverse and positive style, with almost a kind of “dip” back towards negativity in their experimental album.

Popularity scores for all of BMTH’s albums


This plot shows the popularity scores for the songs on all of Bring Me The Horizon’s albums, they are in order of their release date from oldest to most recent. The bigger transparent red/pink circles show the mean popularity for each album. It is easy to see that their older music is the least popular of all and That’s The Spirit and POST HUMAN: SURVIVAL HORROR are their most popular albums. Considering the fact that their music style really started changing on Sempiternal, it is not weird to see that the popularity shoots upwards from that album on. Of course, the experimental album Music to listen to~ creates a big dip in popularity, since it is completely different from what the band usually produces and is not widely liked by fans.

Chroma

Chromagram for the saddest song in Bring Me The Horizon’s repertoire


After plotting the energy and valence of Bring Me The Horizon’s songs over the years, it gave me some insight into which songs were on the outer ends of the ‘emotional spectrum’. Because of this I decided to take a closer look at the happiest and the saddest songs in their repertoire.

This is a chromagram of the song Memorial, which is, according to it’s measurements of valence and energy, the saddest song in Bring Me The Horizon’s entire repertoire. As you can see, the song is quite repetitive and has a few layers that really stand out. Memorial is an entirely instrumental song with a continuous underlying sound and is used in the album as a transition into the next track. There are clearly audible, and in the chromagram visible, high and loud ‘screeching’ noises in the song and a few notes that are softly and repeatedly being played throughout the song. The texture of the song is very simple and also very clearly repesented in this chromagram.

Chromagram for the happiest song in Bring Me The Horizon’s repertoire


This is a chromagram of the song 1x1 (feat. Nova Twins), which is, according to it’s measurements of valence and energy, the happiest song in Bring Me The Horizon’s entire repertoire. As you can see, this song is all over the place. The line of C#/Db stands out a little bit, with a higher magnitude overall. The song consists of heavy drums and guitar combined with electronic music and keyboard sounds, all accompanied by the singing voices of either Bring Me The Horizon’s frontman Oliver Sykes or the featuring artist the Nova Twins. The only big visible thing that is happening here is at around 160 seconds, where the heavy instruments that support the rest of the song go quiet and we hear only a keyboard and a singing voice.

Timbre

Cepstrograms for the most “average” songs


These are cepstrograms for the songs that lay the closes to the average energy, valence and danceability in BMTH’s old music (Sleep With One Eye Open) and new music (Ouch). Cepstrograms show how the distribution of timbre in the music. As you can see the cepstrograms both have more intensity in the bottom, but Sleep With One Eye Open is “squished down” even more than Ouch. Both songs have a focus on the c02 section and generally go up to c06. Even though both songs do not have a lot of intensity in c07 to 12, Ouch does seem to have a bit more intensity there than Sleep With One Eye Open. This shows that the timbre features of the songs are quite different from each other.

Timbre changes in categories


Need to add information

Pitch & chords

Chordograms for BMTH’s singles


These are chordograms for some of the singles of all of Bring Me The Horizons albums in the old and new music categories (the experimental catagory doesn’t have any singles), arranged from oldest to newest (left to right, top to bottom). As you can see, there is a lot going on in all these songs. When listening to the music you can easily follow along in the chordograms. Loud percussion and distorted guitar sounds can make for bright yellow lines, like the one around 70 seconds in Shadow Moses. Considering this, it is clearly visible that Bring Me The Horizon keep using this a lot in their music over time. It is interesting to see that the screaming in Pray For Plagues leads to less intense recognition of chords, but does not necessarily create the bright yellow areas mentioned before.

When comparing the songs to see what happened to BMTH’s music over time, you could say that their music has become more varied. The older songs have a simpler pattern than the newer songs, which have a lot of changes in them.

Keys & modes

Changes in modes throughout Bring Me The Horizon’s albums


In their first two albums, the music is mostly in a major mode. The proportions suddenly flip around for There is a Hell…, seem so even out a bit again for Sempiternal, but the minor mode strongly overpowers again in their next couple albums (though it seems that there are more and more major songs coming with each album). Unlike a lot of other data shows, there are no great differences in Music to listen to~. POST HUMAN: SURVIVAL HORROR seems to go back towards the ratios of Sempiternal a bit more, although mirrored.

How keys and modes change throughout Bring Me The Horizon’s music


This graph shows the keys and modes for the different categories I have divided Bring Me The Horizon’s music into. It is remarkable, that - even though their new music can generally be classified as happier and more positive, as we’ve seen earlier - we see more minor modes in that category than in their old, sadder and angrier music. Their least favourite key seems to be D# (in either mode really) and most of their songs seem to be in C, C# and G for both their old and new music. Their old music also focuses a bit more on G# major. For their experimental music, however, they focus on a key that is not very common in the rest of their repertoire, which is D. Mostly D major, while most of that music is in the minor mode. The only thing that can not be derived from Spotify’s information, is whether they use sharps (#) or flats (♭).

Tempo

Tempo analysis for Bring Me The Horizon’s repertoire


This plot displays the tempo in beats per minute for Bring Me The Horizon’s entire repertoire, the albums are sorted by their release date and each dot represents a song. The bigger red/pink circles represent the mean tempo for each album. It is interesting to see that their old music has the lowest tempo and it starts climbing up when their music starts to drift away from the “metalcore” genre. In their newer music the mean line almost seems to be coherent with the means of popularity for these albums.

Classification

Classification


Here I have trained a classifier to distinguish between the three different categories I divided Bring Me The Horizon’s music into. After trying different combinations, I figured out that energy, valence, duration and all of the timbre components are the most useful features for computational recognition of this corpus. The keys generally confused the classifier the most and duration seemed to influence the experimental category the most (significantly fewer songs were classified correctly). As you can see in the mosaic plot, most of the music is classified in the right category. It is important to mention that for both old and new 30 songs were used, but there are only eight experimental songs. Considering this, it’s quite remarkable that five out of those eight are classified correctly here.

The precision rates are very high for all of the categories, 83% for experimental music, 82% for new music and even 96% for old music. The categories I divided the music into are thus very distinguishable from each other, especially when filtering out the right elements (even without doing this the scores are still relatively high).

Conclusion

Coming soon…